Pisa_results <- read.csv("Pisa_results.csv")

What is PISA?

PISA (Programme for International Student Assessment) is an international survey that aims to compare the performance of students at the age of 15 from different countries in science, mathematics and reading. The PISA test is conducted every three years, but we will only focus on the results of the 2018 test. The PISA test aims to help improve the school systems around the world.

The score scale is specifically set to determine the average level of education in a country, while the resulting score is an average of all students’ scores. The evaluation of PISA results makes it possible to identify the weaknesses of educational systems and take steps to improve them.

What we focused on

We analysed data frames showing the results of PISA in mathematics and science. We focused on the average score in the tests by country in this fields of study.

Plot

The graph below shows the relationship between science and maths scores. We have grouped the countries by the continents they are located on. By hovering over the individual dots you will see the name of the country and its scores.

Pisa_results <- Pisa_results %>% 
  filter(is.na(Continent) == FALSE, Country != "Viet Nam")



Pisa_results$Math <- as.double(Pisa_results$Math)
Pisa_results$Science <- as.double(Pisa_results$Science)

fig <- plot_ly(data = Pisa_results,
               x = ~Math,
               y = ~Science,
               color = ~Continent,
               colors = "Set1",
               type = "scatter",
               text = paste0("Country: ", Pisa_results$Country),
               hovertemplate = paste('<b>%{text}</b><br><b>Mathematics</b>: %{x}<br><b>Science</b>: %{y} <extra>tooltip</extra>')) %>% 
  layout(
    title = "Mean scores from PISA 2018",
    xaxis = list(title = "Mathematics",
                 range = c(300, 600)),
    yaxis = list(title = "Science",
                 range = c(300,600))) 
fig

First observation

A very strong linear relationship can be observed between the results in the graph. However, do the countries of each continent follow the same pattern? When selecting each continent, different correlations can be observed. An interesting example is Asia, where the differences in performance between East and West Asian countries are the most visible.

Second observation

We might wonder what the average results of the countries with the best universities in the world look like. We are interested in the countries with the best universities in the world according to the QS ranking, that is the USA, England, China, Switzerland and Singapore. It is to be expected that these countries should be at the top of the chart. The PISA tests are scored on a scale so that the average score of OECD (Organisation for Economic Cooperation and Development) countries is 500. In the table below we can see that these countries are most often above the average.

Even though the USA, the UK and Switzerland score are around average, they still have the most prestige universities in the world. It is only Singapore and China that are transferring the level of basic education to their universities.

df <- Pisa_results %>% 
  filter(Country %in% c('B-S-J-Z (China)', 'United States*', 'Singapore', 'United Kingdom', 'Switzerland')) %>% 
  select(Country, Math, Science)
df %>% 
  knitr::kable(caption = 'Scores in chosen countries')
Scores in chosen countries
Country Math Science
Switzerland 515 495
United Kingdom 502 505
United States* 478 502
B-S-J-Z (China) 591 590
Singapore 569 551

Conclusions

Among the countries taking part in the PISA survey, a linear relationship can be observed between the results. East Asian and Central European countries are in the lead. In Europe, the Balkan countries are the worst performers, and in Asia there is a gap of almost 80 points between East Asia and the rest.

The graph also helps us to see that the level of higher education and primary education does not have to go hand in hand in countries above the 500-point average.